Sept. 5, 19t52
REACTIONS OF VANILLJN AND ITS
3,4,5-Trimethoxybenzaldehyde.-5-Hydroxyvanillin was methylated with dimethyl sulfate and alkali following the procedure employed by Buck7 for the methylation of vanillin to veratraldehyde, and the product was recrystallized from water to give 77% of pure 3,4,5-trimethoxybenzaldehyde melting a t 74-75' and not depressing a mixed melting point with authentic 3,4,5-trimethoxybenzaldehydeprepared by methylating gallaldehyde according t o Rosenmund and Zetzsche.8 Syringa1dehyde.-A mixture of 10 g. of recrystallized 8,3,5-trimethoxybenzaldehyde and 56 g. of concentrated 5ulfuric acid was maintained a t 40" in a water-bath for 8 hours and then allowed to stand overnight a t room temperature. The mixture was stirred into 100 cc. of cold water and cooled. The clear solution was decanted from the little tar that separated and was extracted with ether. T h e ether mds dried and distilled t o yield 8.9 g. (96%) of almost pure (as indicated b y chromatography) syringaldehyde which, upon crystallization from petroleum ether (b.p. 65-IlO"), melted a t 109" and did not depress the melting point of a mixture with authentic ~yringaldehyde.~ The use of crude 3,4,5-trimethoxybenzaldehyde or of longer reaction times resulted in poorer yields and in much cruder products. Reaction of 5-Hydroxyvanillin with One Mole of Potassium Hvdroxide and One Mole of Dimethyl Sulfate.-5Hydroxsvanillin (16.8 g., 0.1 mole) was treated with 7.7 g. (0.1 mole) of potassium hydroxide in 60 cc. of water and with 12.6 g. (0.1 mole) of dimethyl sulfate in the same manner employed for the complete methylation noted above. The cooled reaction mixture was extracted with ether and the ether was washed with 5 % sodium hydroxide solution and then with water. The alkaline solution and washings were acidified and extracted with ether t o yield 16.0 g. of oily product. The original ether solution, when dried and distilled, yielded 3.3 ,g. of crude 3,4,5-trimethoxybenzaldehyde melting a t 70 . Recrystallization from petroleum ether (b.p. 6 5 1 2 0 " ) yielded crystals melting a t 74-75" which did not depress a mixed melting point with authentic 3,4,5-trimethoxybenzaldehyde. The oily phenolic product was chromatographed from benzene on acid-washed Magnesol and developed with benzene-ethanol (10O:l). The bands were located by streaking with 2,4-dinitrophenylhydrazine, potassium perman(7) J . S. Buck, "Organic Syntheses," Coll. Vol. 11, John Wiley and Sons, Inc., New York, N. Y., 1943, p. 619. (8) K . W. Rosenmund and F. Zetzsche, Bcr., 01, 594 (1918). (9) I. A. Pearl, THISJOURNAL, 70, 1746 (1948).
DERIVED COhlPOUNDS
4263
ganate and ferric chloride.10 The leading band was eluted with acetone to yield 40% of 5-hydroxyveratraldehyde melting a t 63-64" and not depressing a mixed melting point with authentic 5-hydro~yveratra1dehyde.l~T h e next band, upon elution, yielded 13% syringaldehyde melting a t 109l l O o , and the upper band yieldedo 20% of unchanged 5hydroxyvanillin melting a t 132-133 Reaction of 5-Bromovanillin with Sodium Hydroxide Solution in the Presence of Active Copper.-Active copper powder was prepared according t o Brewster and Greening.'? A mixture of 155 g. of 5-bromovanillin, 50 g. of freshly prepared active copper powder and 3000 cc. of 8% sodium hydroxide solution was placed in a one-gallon stirring autoclave, heated t o 200-210" with stirring for one hour and allowed to cool with stirring. The copper was filtered, and the alkaline solution was acidified with dilute sulfuric acid. Some tar was removed by filtration, and the filtrate was concentrated to a small volume in a vacuum circulating evaporator below 50". T h e tar was boiled with water and filtered hot. Cooling of this filtrate yielded 4.4 g. of crude 5-bromovanillic acid which, recrystallized from water in the presence of charcoal, yielded white needles melting a t 226' and not depressing a mixed melting point with authentic 5bromovanillic acid.IJ The concentrated original aqueous filtrate deposited a crystalline precipitate on cooling. The crystals were filtered and recrystallized from water to yield 5.1 g. of vanillic acid melting a t 210-211'. T h e filtrate was extracted with ether, and the ether was extracted successively with 21% sodium bisulfite, 8% sodium bicarbonate and 5% sodium hydroxide solutions. Work u p of the bisulfite extract and chromatographing the residue yielded 6.6 g. of vanilliii and 1.8 g. of 5-hydroxyvanillin. The bicarbonate solution yielded 0.8 g. of vanillic acid. The sodium hydroxide extract, upon acidification and extraction with ether, yielded 13.6 g. of phenolic oil which was distilled a t 0.4 mm. pressure t o give pure guaiacol, whose benzoate melted a t 58-59" and did not depress a mixed melting point n ith the benzoate prepared from authentic guaiacol. Guaiacol was also recovered from the condensate of the original vacuum concentration.
.
(IO) I. A. Pearl and E. E. Dickey, ibid., 73%863 (3953). (11) E. Spath and H. Roder, Monafsh., 43, 106 (1922). (12) R. Q. Brewster and T. Groening, "Organic Syntheses," Coil. Vol. 11, John Wiley and Sons,Inc., New York, N. Y., 1943, p 446. (13) L. C . Raiford a n d D. J. Potter, THIS JOCRKAI., 06, 1682 (1033).
APPLETOK, \YISCOSSIN
[CONTRIBUTION FROM THEINSTITUTE O F
PAPER CHEMISTRY]
Reactions of Vanillin and its Derived Compounds. XVIII. Preparation and Reactions of Lignin Model Substances Related to 5-Carboxyvanillic A ~ i d ~ , ~ BY IRWIN A. PEARL AND DONALD L. BEYER RECEIVEDMARCH 5, 1952 5-Allylvanillin was isomerized by low temperature caustic fusion i o yield 5-propenylvanillin. Ethyl vanillate and ethyl orthovanillate were converted t o ethyl 5-allylvanillate and ethyl 5-allylorthovanillate, respectively, and these were treated directly with fused caustic to yield 5-propenylvanillic acid and 5-propenylorthovanillic acid. Oxidation of the propenyl compounds with nitrobenzene and alkali a t 160-165" gave the corresponding formyl derivatives as the major products of reaction along with other intermediates. All the formyl derivatives were oxidized by means of alkali and silver oxide to 5-carboxyvanillic acid.
The recent isolations of 5-carboxyvanillic acid and 5-carboxyvanillin from metallic oxide oxida( I ) For Paper XVII of this series, see THISJOURNAL,74, 4262 (1952). ( 2 ) Presented before the Division of Cellulose Chemistry a t t h e l 2 l s t Meeting of the American Chemical Society, Milwaukee, Wisconsin, March 30 t o April 3, 1952. (3) This paper represents a portion of the results obtained in t h e research program sponsored by t h e Sulphite Pulp Manufacturers' Research League and conducted for t h e League by T h e Institute of Paper Chemistry. Acknowledgment is made by t h e Institute for perniission o n t h e part of the League t o puhlish these results.
tions of lignosulfonate r n a t e r i a l ~ ~led - ~to a study of the preparation of these materials and of the reactions of the lignin model substances employed in their preparation. The present paper describes the preparation of the propenyl derivatives of. vanillin, vanillic acid and orthovanillic acid and their oxidation with nitrobenzene and alkali to form 5-formylvanillir1, 5-formylvanillic acid and 5(4) I A Pearl, THIS JOURNAL, 72, 1427 (1950) ( 5 ) I. A Pearl, ibid , 72, 2309 (1950) (6) I. A Pearl and D L Reyer, T o p p r , 33, 508 (1030)
IioL 100
n
90
5
80
.-
70
0 .* + i
A
60
%
x
r/:
1,,UL
50
40
20
10 0
200
240
2 80
320
360
400
U7nveleiigth, tnp.
Fig. 1.--Ultraviolet absorptioii spectra in 95', ethxnol: L.11, 5-allylvariillin; Y I I I , 5-propeiiylvanilliii (ri.7 :md f r i i n s : ; I S , ci-forniSlv,Lnilliii.
Fig. 2.-Ultraviolet absorption in 95% ethanol: 11, ethyl vanillate allyl ether; I V , 5-allylvatiillic acid ; 17, 5-propenylvanillic acid.
carboxyvariillin, respectively. The latter coinpounds are all oxidized by alkali and sil\rer oxide to yield 5-carboxyvariillic acid. Attempts to further isomerize 2-allylvanillin (VII)' to 5-propeiiylvanillil1-lvanilliii (VIII) with potassium hydroside in amyl alcohol by the general procedure of Tiemann* or with potassiuin hydroxide in inethan01 by the general procedure of Klagesg gave only sinall yields of the desired compound. 5-Fropenylvanillin was easily prepared in good yield by the fusion of 3-allylvanillin with potassium hydroxide a t 100-150' arid was obtained as white needles melting a t i s 3 . Pew'O has very recently reported the preparation of 5-propenylvanillin melting at 98-99' by the treatment of 5-allylvanillin with potassium hydroxide in methanol. The 73' and the 98" compounds are apparently cis-trans isomers. The cis-73' conipound is easily converted to the trans-98' compound by boiling with methariolic potassiuni hydroxide, but the truns-9s' product is stable under conditions originally forfiii ng the cis-compound. Ethyl vanillate (I) was converted to ethyl 3allylvanillnte (111) in the same manner, and the resulting ester was hydrolyzed to 5-allylvanillic acid (IV) melting a t 147-148'. By essentially the same reactions Takahashi and Kawashiina" obtained the same intermediates, but recorded 200202' as the melting point for 5-allylvanillic acid. Low temperature caustic fusion of either 5-allylvanillic acid or its ethyl ester resulted i n the formntioii of 5-propenylvanillic acid (V). Similarly,
ethyl orthovanillate (XI) was transformed into ethyl 5dlylorthovanillate (XIII) which was isomerized directly by fusion with potassium hydroxide to 5-propenylorthovanillic acid (XIV), a compound previously prepared in an analogous mariner from methyl orthovanillate by Freudenberg and Klink.12 Preliminary oxidation of the above propenyl compounds by boiling with nitrobenzene and alkali a t atmospheric pressure under conditions employed for the conversion of isoeugenol to vanillin13resulted in recovery of only the starting material in the case of 5-propenylvanillin and in the formation of only sinall yields of the desired formyl derivatives in the cases of the other two propenyl compounds. The propenyl compounds were then treated with alkali arid nitrobenzene a t superatmospheric pressure under conditions similar to those employed by Freudenberg, Lautsch and Engler14 in their original study on the oxidation of lignin materials with nitrobenzene and alkali. Such oxidation of cis5-propenylvanillin gave a good yield of 5-formylvanillin (Ixj and 5-carbosyvanillin (XV) and a sinall amount of 5-carboxyvanillic acid (X) as a by-product. 5-Formylvanillin was first prepared by Koetschet and Koetschet15 as a by-product in the preparation of vanillin by the Reimer-Tiemann reaction on guaiacol. The high yield of 5-carboxyvanillin in this experiment and the absence of 5forinylvanillic acid indicate that the formyl group ortho to the hydroxyl group in the guaiacyl nucleus is more susceptible to oxidation than the formyl group in the para position. This fact helps to explain the presence of only vaiiillin deriv a t'ives in lignin oxidation mixtures and the apparent at)-
( 7 ) I.. Cloisen a n d 0 . E i 4 e b , A ~ t n .401, , 112 ( l ! l l 2 ) ( 8 ) 17, Tiemann, Ber., 24, 2871 (1891). (9) A Klnges, ibid , 39,2591 (1906). ' 111 J C I ' r i r , 'l'irrs J O I . R S % I . , 73, Ifis-l fl',l:l) i ! ) 'I' 'l':+!.aha l ~ ilr,tract oiily a v a i l j b l c
(12) K. Freudenberg arid P. Klink, Ber., 73, 1 3 7 1 (19.10). (13) J. 1). Riedel, Ijritisli Patent 253,451 ( 8 e p t 1:3, lP28). (11) K , l ~ r e t ~ c l c n l i w g\L', T.u,ltwli niirl K . l < ~ i g l c ~Ilei. r , , 73, 171 (l!14(Jl, ( 1 5 ) J Kiwlwhrt a n d P. Gor1sc:tirf. I l ~ l a . Chim . J r l ~ .13, 452 ~lmo).
REACTIONS OF VANILLINA N D
Sept. 5, 1952
ITS
DERIVED COMPOUNDS
4265
10090
-
80 70
s
.-e
.*: i
-
6050-
U
5
403020
-
10
-
%% 200
240
280
320
380
400
Wave length, mp. Fig. 3.-Ultraviolet absorption spectra in 95y0 ethanol: X I I , ethyl orthovanillate allyl ether; X I I I , ethyl 5-allylorthovanillate; XIV, 5-propenylorthovanillic acid.
, '\
240
280
320
4
360
400
Wave length, mp. Fig. 4.-Ultraviolet absorption spectra in 95% ethanol : VI, 5-formylvanillic acid; X, 5-carboxyvanillic acid; XV, 5-carboxyvanillin. .
Ultraviolet absorption spectra of the above compounds were determined in 95% ethanol with a Beckman spectrophotometer a t minimum slit width. These spectra are pictured in the figures. sence of orthovanillin derivatives. Under the The effect of isomerization of allyl ethers to carbonsame conditions trans-5-propenylvanillin gave a allyl linkages is very apparent in all cases as are small yield of the desired 5-formylvanillin and a the isomerizations from allyl side chains to prohigh yield of unchanged starting material. Thus, penyl side chains. Ultraviolet absorption spectra the cis-compound appears to be much more sus- serve admirably for distinguishing these isomers and ceptible to nitrobenzene oxidization. proved very useful in following the progress of some Similar oxidation of 5-propenylvanillic acid gave of the experiments in this work. The ultraviolet 5-forrnylvanillic acid (VI) and 5-carboxyvanillic absorption curves determined on these synthetic acid as the only oxidation products, both in fair compounds have proven invaluable in the iden tiyield. On the other hand, oxidation of 5-propenyl- fication work we have carried on with the products orthovanillic acid yielded over 90% of 5-carboxy- isolated from lignin oxidation reactions. vanillin and only a trace of 5-carboxyvanillic acid. Again, these data appear to indicate that the Experimental formyl group ortho to the hydroxyl group in the All melting points are uncorrected. guaiacyl nucleus is more susceptible to oxidation 5-Propenylvanillin (VIII).-A mixture of 500 g. of potassium hydroxide and 150 cc. of water was heated by means of than the formyl group in the para position. It is interesting to note that neither vanillin nor an electric hot-plate in a stainless steel beaker t o 140' and treated with vigorous stirring with 200 g. of 5-allyl~anillin.~ vanillic acid were produced in any of the above The mixture became thick and orange colored. The temsuperatmospheric oxidations with nitrobenzene and p x a t u r e was raised t o 160", at which point t h e hot alkali through decarboxylation. Leopold16oxidized plate was turned off. The temperature continued to rise t o 5-allylvanillin (VII) with nitrobenzene and alkali 180", and the mixture became quite fluid. Stirring was continued, and the temperature remained above 160" for and obtained a negligible amount of vanillin, 20% about 15 minutes. When the temperature reached 120', 5-formylvanillin and 6% 5-carboxyvanillin. the mixture was diluted with 4 liters of water t o effect solOxidation of either 5-formylvanillic acid or 5- ution of the fused mass and was stirred until cool. The carboxyvanillin with one mole of silver oxide and cold alkaline solution was acidified with dilute sulfuric acid, the oil which separated solidified upon stirring and excess alkali by a previously described p r o c e d ~ r e ' ~ atld cooling. The yield of crude 5-propenylvanillin was 200 gave substantially quantitative yields of 5-car- g. or 100%. Recrystallization from either water or petrolboxyvanillic acid. Analogous oxidation of 5- eum ether gave white needles melting at 73'. Anal. Calcd. for CIIH~ZOB: C, 68.74; H, 6.29; CHaO, formylvanillin with two moles of silver oxide and excess alkali yielded 5-carboxyvanillic acid, along 16.14. Found: C, 68.88; H, 6.40; CHBO,16.14. Ethyl Vanillate Allyl Ether (II).-This compound was with 5-carboxyvanillin and some resinous products. prepared by allylation of ethyl vanillate and was obtained Perkins and Stoyle's noted that 5-carboxyvanillin as a colorless oil boiling at 144-143' at 1.4-1.5 mm., nZ6D reduced ammoniacal silver nitrate but they did not 1.5370. Takahashi and Kawashimall prepared this ester and reported a pale yellow oil boiling a t 160" at 10 mm. record the fate of the 5-carboxyvanillin. Vanillic Acid Allyl Ether.-The allyl ether of ethyl vanil(IG) B. Leopold, Aria Chcm. Scand., 4, 1523 (1950). (17) I. A. Pearl and D. L. Beyer, THIS J O U R N A L , 74, 3188 (1952). (18) W. H. Perkins and F. W. Stoyle, J . Chem. Sac., 128, 3171 (1923).
late was boiled with a dilute alcoholic solution of sodium hydroxide, and the clear solution obtained was cooled, filtered, diluted with water and acidified with dilute hydrochloric acid. The heavy precipitate was filtered, washed
with water, aiid recrystallized from dilute ethanol in the presence of carbon t o yield white crystals inclting a t 175-
colorless crystals of 5-forinylvaiiillic acid melting a t 222223" and giving a deep violet color with ferric chloride soluIi~i'. tion. The yield was 5.5 g . . l ~ t d . Calcd. for CLIHIiOd: C , 63.45; 13, 3.81. Found: Anal. Calcd. for CdHa05: C, 55.11; H, 4.11; CIIIO, C, 63.46; H, 5.93. 13.82. Found: C, 55.22; H, 4.19; CHnO. 15.83. Tiemann and Mendelsohnl@prepared 5-formylvanillic Ethyl 5-Allylvanillate (111).--This compound was prepared by rearrangement of ethyl vanillate allyl ether accord- acid by the Reimer-Tiemann reaction on vanillic acid and ing to Takahashi and Kawashima" and was obtained as reported a melting point of 221-222". Similar treatment of 5-propenylorthovanillic acid with iridescent plates from petroleum ether ( b . p . 30-60") meltnitrobenzene and alkali at atmospheric pressure resulted in ing a t 19-50". 5-Allylvanillic Acid (IV) .--Ethyl 5-allylvanillate was the recovery of most of the starting material. In addition, warmed on the steam-bath with dilute aqueous sodium hy- 4.2 g. of 5-carboxyvanillin melting a t 256-257" (from 70 g. droxide. After 30 minutes the clear solution was acidified of 5-propenylorthovanillic acid) was obtained from thc sowith sulfur dioxide. The white precipitate was filtered and dium bisulfite-soluble fraction. Analogous treatment of S-propenylvanillin yielded only rccrystallized from water t o give fine white needles melting the starting material. .it 147-148'. 5-Carboxyvanillin (XV).-A mixture of 65 g. of 5-propenyl.117ai. Calcd. for C I I H l . ~ O aC, : 63.45; H, 5.81. Found: orthovanillic acid, 217 g. of nitrobenzene, 87 g. of sodium C, fK3.60; H, 5.90. hydroxide and 870 g. of water was heated in a one-gallon Takahashi and Kawashimall reported a melting point of stirring autoclave during the course of 0.5 hr. t o 160", main200--202° for 5-allylvanillic acid. tained a t 160 f 5' for one hour, arid allowed to cool. The 5-Propenylvanillic Acid (Vj.-.4 mixture of 100 g. of po- mixture was diluted with water and extracted conipletely tassiurn hydroxide and 50 g. of water was heated with stir- with ether to remove all nitrobenzene and its reduction riiig to 120' and treated with 45 g. of 5-allylvanillic acid. products. The aqueous solution was acidified with sulfur The reaction mixture became semi-solid. The temperature dioxide and extracted with ether. The ether was dried arid \vas raised to 160° and maintained a t 160-180' for 10 distilled to yield 4.5 g. of crystalline residue. Leaching with minutes. After cooling, the caustic mixture was diluted 21 % sodium bisulfite solution left a little insoluble material with water and acidified with sulfur dioxide. The crude 5- which, upon recrystallization from water, gave a few mg. of propenylvanillic acid weighed 42 g. (93%). Recrystalliza5-carboxyvanillic acid melting a t 271-272". Acidification tion from dilute ethanol, dilute methanol, or benzenr- of the bisulfite solution and extraction with ether yielded petroleum ether yielded white crystals melting a t 1 2 5 1 2 6 " . 4.2 g. of 5-carboxyvanillin melting nt 256-258". Anal. Calcd. for CIIH1~O4: C , 63.45; H, 5.81. Found: The original sulfur dioxide-saturated aqueous solutioii C, 63.41; H , 5.88. was acidified u-ith sulfuric acid and boiled to remove dis5-Propeiiylvanillic acid was also prepared directly from solved sulfur dioxide. A heavy crystalline precipitate separated in the boiling solution. The mixture was alloivcd to te (111). A fusion mixture of 400 g. of de and 200 cc. of water a t llU0 was cool and was then filtered. The precipitate was washeti g. of ethyl 5-allylvanillate. A viscous with water and dried to give 45 g. of crude 5-carboxyvanillin sludge of the potassium salt of the ester resulted. The tern- which, upon recrystallization from water, melted a t 257258'. The filtrate from this precipitate was extracted with pcrature was raised with continued stirring, and at 120-130' ether, and the ether was dried and distilled to yield 5.5 g. ;t vigorous reaction took place, resulting in a more viscous slurry. The temperature was gradually raised to ZOO", of crude 5-carboxyvanillin which also yielded crystals melting a t 256-257' upon recrystallization from water. Thc aiid then the mixture was allowed t o cool. Slight reaction took place all the time until 200" was reached. When the total yield of crude .i-carboxyvanillin was 90%. A mixed melting point of this 5-carboxyvanillin with that mixture had cooled to 120", i t was diluted with 2.5 liters of water and stirred to effect solution. The clear solution was isolated from a cupric oxide-alkali oxidation of spent sulfite .icidified with sulfur dioxide, and the precipitate was filtered liquor5 gave no depression. 5-Formylvanillic Acid (VI! .-5-Propenylvanillic acid iv:t\ : i i i t l dried t o yield 145 g. (82%) of crude 5-propenylvanillic ,icitl which wits recrystallized from dilute methanol to give t I cated with nitrobenzene anti alkali in the identical ~ n n i i i i e r , cxcept that the reactioii misture was maintained a t l f i 0 3 \vliile crystals melting a t 125-126'. '. Trc;itmeiit of the cther-cxtracted alkalitic Ethyl 5-Allylorthovanillate (XIII).-Ethyl orthol-anillatc -lure with iulfur dioxide precipitated sonic uiii S I ! WAS allylated in a manner identical with that employed acid. Ether extraction of the fc>rthe prcpar:ltiou of the allyl ether of ethyl vanillate. The rcactctl ~j-]~roi)enyl\~aiiillic filtered sulfur cliosidc-satut-;itetl solutioii yielded 16 g. of 3cillyl c:lher of ethyl orthovanillate ( M I ) was rearrangctl carboxyvanillic acid i S I , incllirig a t 2'71--272'. Repeated directly by boiling under reflux for one hour at 60 mm. prcsiure and then distilling under vacuum t o yield XI11 as a recrystallization from water vielded crystals melting a t 281 s viscous oil boiling at 128-132" a t 0.5 nun., n z o ~ 282". The sulfur cliosiclc-saturated aqueous solution was acidified with sulfuric acid, boiled t o expel sulfur dioxide, and The over-all yield was 88%. .-1naf. Calcd. for C13H1604: C, 66.08; H, 6.83. Found: cooled. A light tan crystalline solid separated. This was filtered, washed with water, and dried t o give 39 g. of C. 66.06; H. 6.83. 5-formylvanillic acid which, upon recrystallization from 5-Propenylorthovanillic Acid (XIV).-Treatment of ethyl dilute ethanol, water, or dioxanepetroleum ether, yielded 3-allylorthovanillate with fused caustic by the same proce- colorless crystals melting a t 223--224'. A mixed melting dure employed for the preparation of 5-propenylvanillic point with authentic 5-forniylvanillic acid was not depressed. acid from ethyl 5-allylvanillate gave a n 89% yield of 55-Formylvanillin (IX).-Similar treatment of '75 g. of cis~)ropenylorthovanillic acid which, upon recrystallization 5-propenylvanillin with 230 g. of nitrobenzene, 100 g. of from water, yielded colorless crystals melting a t 155". sodium hydroxide and 1000 g . of water a t 165' for otic hour Fi-cutlcnbcrg and KlinkI2 reported 157" as the melting point gave a reaction mixture which was diluted, cooled, and exof X I V prepared by caustic fusion of methyl %5-allylortho- tracted with ether. The aqueous alkaline solution was vanillate a t 235'. acidified with sulfur dioxide, cooled, and extracted with Oxidations with Nitrobenzene and Alkali a t Atmospheric ether; the ether was dried and distilled to yield 5.2 g. of Pressure.-A mixture of 35 g. of 5-propenylvanillic acid, 140 crude 5-carboxyvanillic acid which was recrystallized from g . of nitrobenzene, 50 g. of sodium hydroxide and 150 cc. of water to give colorless crystals melting at 28G28I0. Thc water was boiled under reflux for 6 hours, cooled, diluted sulfur dioxide-saturated solution was acidified with sulfuric \rith water, and extracted with ether. The alkaline aqueous acid and boiled. The solution was cooled and extracted solution was acidified with dilute sulfuric acid. A heavy with ether. The ether was dried atid distilled t o yield 44.0 S ITpon crystallization from dilute precipitate H - ~ obtained. g. of .F-formylvanillin which, upon recrystallization from ethanol or water, the precipitate yielded colorless crystals of petroleum ether (b.p. & - - I l o o ) , gave colorless spherical the 5t:rrtiiig .j-propeiiylvaiiillic acid. The filtrate from this crystals melting a t 125 -126'. Koetschet arid Koetschet l5 precipitate was extracted with ether, and the ether was ex- recorded I19---121O as the melting point of 5-forinylvanillin t r z t e t l with 21 % sodium bisulfite solution. The bisulfite prepared as a. by-product in the preparxtion of vanillin and e x t r a c t was acidified with sulfuric acid, boiled, cooled and orthovanillin h y the Rcitiier-Tieri~;tnn procc>>. ... . .. . cz~r-actctl with ctlicr. The cthcr \vas tlricd a i i d distillctl, , ~ : i ( i: h e i-ciidiic: T\:LS i-ccryit;illizctl t u icc Twin ~ r a t c rto g i \ c
